LLM-evolved game levels: one global function takes Zelda playability from ~0% to ~100%
Amidos2006 · x · 2026-09-09
- Method: use an LLM (Claude Opus 4.8) as the code generator, representing all level rules as Python code; the prompt asks for new useful functions that follow existing signatures and differ from the one used by the current chromosome.
- Key result: every problem except Binary needs at least 1 global function to become highly playable — Zelda goes from 0% playability with 0 globals to 100% with 1 global.
More from Research
- If human insight remains an ineliminable complement, it will limit AI's gains in science — erikphoel · 2026-09-10
- Simplex launches research blog Belief Updates, showing transformer belief geometry forms telescoping cones — rishabh16_ · 2026-09-10
- After Navier-Stokes: the anti-hype take on AI's role in the math breakthrough — Jsevillamol · 2026-09-10
- DEER-3D fixes 3D-LLM grounding biases with error-driven counterfactual scene editing — YonatanBitton · 2026-09-10
- Universities' official policies contain straight-up misinformation about AI detectors — paulnovosad · 2026-09-10
- LangChain's Chase: Agent Gains Come from Harness, Not Model — Paper Shows 14pt Lift — hwchase17 · 2026-09-10